Time course of bilateral microglial activation in a mouse model of laser-induced glaucoma

Microglial activation is associated with glaucoma. In the model of unilateral laser-induced ocular hypertension (OHT), the time point at which the inflammatory process peaks remains unknown. Different time points (1, 3, 5, 8, and 15 d) were compared to analyze signs of microglial activation both in OHT and contralateral eyes. In both eyes, microglial activation was detected in all retinal layers at all time points analyzed, including: i) increase in the cell number in the outer segment photoreceptor layer and plexiform layers (only in OHT eyes) from 3 d onward; ii) increase in soma size from 1 d onward; iii) retraction of the processes from 1 d in OHT eyes and 3 d in contralateral eyes; iv) increase in the area of the retina occupied by Iba-1+ cells in the nerve fiber layer/ganglion cell layer from 1 d onward; v) increase in the number of vertical processes from 1 d in contralateral eyes and 3 d in OHT eyes. In OHT eyes at 24 h and 15 d, most Iba-1+ cells were P2RY12+ and were down-regulated at 3 and 5 d. In both eyes, microglial activation was stronger at 3 and 5 d (inflammation peaked in this model). These time points could be useful to identify factors implicated in the inflammatory process.

The neurodegenerative disease glaucoma is one of the main causes of irreversible blindness 1 . Standard treatments aim to control intraocular pressure (IOP), but this is often insufficient to prevent loss of visual field 2 . Many studies have attempted to clarify the pathogenic mechanisms in glaucoma, but it remains poorly understood. This multifactorial optical neuropathy is known to involve the progressive loss of retinal ganglion cells 3 .
Microglial cells are activated by an increase in IOP 16 , which causes their soma to enlarge and processes to retract. Activated microglia become migratory and begin to function as antigen-presenting cells and to proliferate 17 . These microglia express and secrete a range of inflammatory mediators into the retina and aqueous humor, including receptors, proteases, reactive oxygen species, cytokines and nitric oxide [18][19][20][21] . Activated microglia cells densely extend their processes toward the lesion site within a few minutes 22,23 , or they migrate to the site of damage 24,25 to prevent lesion spread. These first responses are mediated by metabotropic purinergic P2Y12 receptors, which are stimulated by ATP released from dead or injured neurons 22,23 . P2RY12 is a specific marker for rodent resident microglial cells 26 and is not expressed by monocytes or macrophages 27,28 . This chemotactic receptor regulates microglial activation 29 , and its signaling activity may enable microglia to respond to altered synaptic activity due to various types of injury 29 .
While activated microglial appear to contribute to neuronal death in glaucoma, they also express neuroprotective molecules and help remove debris created as retinal ganglion cells die 30 . Therefore, microglial activation  Fig. 2a,d,e). The variation in OS of OHT and contralateral eyes over time showed that the number of Iba-1+ cells in OHT eyes was significantly higher at 5 d than at 1 d (p < 0.01) or 3 d (p < 0.05), while the numbers at 8 and 15 d were significantly lower than the peak at 5 d (both p < 0.01; Fig. 2a,e). The peak number of Iba-1+ cells in contralateral eyes occurred at 3 d (p < 0.01 vs 1 day; Fig. 2a,e), after which the number decreased significantly by 15 d (p < 0.01 vs 3 d), becoming similar to the number in naïve eyes (Fig. 2a,d,e).
In OHT eyes, no differences were found in the numbers of Iba-1+ cells among the superior, inferior, nasal and temporal retinal zones, either when the analysis was performed over all retinal areas or over the areas nearest to the optic disc.  In contralateral eyes, the cell body area was greater at 3 d, and then it decreased such that it was significantly smaller at 15 d than at 3 d. The dotted line indicates the mean values in naïve eyes. *p < 0.05, **p < 0.01. naïve eyes (OPL and IPL). OHT and contralateral eyes showed several significant differences from naïve eyes in the outer plexiform layer (OPL) and inner plexiform layer (IPL). The number of Iba-1+ cells was significantly higher in OHT eyes than in naïve eyes at 3, 5, and 15 d in both OPL and IPL (all p < 0.01; Supplementary  Fig. S1b,c,e; Fig. 2b-e) and at 8 d in IPL (p < 0.01; Supplementary Fig. S1d; Fig. 2c-e). The number was significantly higher in contralateral eyes than in naïve eyes only at 8 d in IPL (p < 0.05; Supplementary Fig. S1d; Fig. 2c-e). The number of Iba-1+ cells in the OPL was significantly higher in OHT eyes than in contralateral eyes at 3 d (p < 0.05; Supplementary Fig. S1b; Fig. 2b,d,e), 5 d (p < 0.05; Supplementary Fig. S1c; Fig. 2b,d,e) and 15 d (p < 0.01; Supplementary Fig. S1e; Fig. 2b,d,e). A similar result was observed in the IPL at 3, 5 and 8 d (all p < 0.05; Supplementary Fig. S1b,c,d; Fig. 2c-e) as well as 15 d (p < 0.01; Supplementary Fig. S1e; Fig. 2c-e). Soma size was significantly greater in OHT and contralateral eyes than in naïve eyes at all time points (all p < 0.01; Supplementary Fig. S1f-j; Fig. 2d-g).
In OHT eyes, no differences were found in the numbers of Iba-1+ cells among the superior, inferior, nasal and temporal retinal zones in the OPL or IPL, either when the analysis was performed over all retinal areas or over the areas nearest to the optic disc.
At all time points, the arbor area of Iba-1+ cells in the OPL was significantly smaller in OHT eyes than in naïve eyes at 1 d (p < 0.05; Supplementary Fig. S2a; Fig. 3a,e,f) as well as at 3, 5, 8 and 15 d (all p < 0.01; Supplementary  Fig. S2b-e; Fig. 3a,e,f), indicating a retraction of processes. Similarly, the area in IPL was significantly smaller in OHT eyes than in naïve eyes at 1, 3, 5, 8 and 15 d (all p < 0.01; Supplementary Fig. S2a-e; Fig. 3b,e,f). These results with OHT eyes were mirrored in those with contralateral eyes, in which the arbor area in OPL was significantly smaller than in naïve eyes at 3, 5 and 8 d (all p < 0.01; Supplementary Fig. S2b-d; Fig. 2a,e,f) as well as 15 d (p < 0.05; Supplementary Fig. S2e; Fig. 3a,e,f). The arbor area in IPL was significantly smaller in contralateral eyes than in naïve eyes at 3, 5,  Fig. 3c). At 1 d, the number of processes was significantly greater in contralateral than OHT eyes (p < 0.05; Supplementary Fig. S2f; Fig. 3c). However, at 15 d, the number of processes in OHT eyes was significantly greater than in contralateral eyes (p < 0.01; Supplementary Fig. S2j; Fig. 3c).

Variation in plexiform layers of OHT and contralateral eyes over time. Number of Iba-1+ cells in OPL and IPL.
In OHT eyes, the largest increase in the number of Iba-1+ cells from 1 d occurred at 3 and 5 d (both p < 0.01; Fig. 2b,c,e). The number decreased from 5 d until 8 d and 15 d (both p < 0.01; Fig. 2b,c,e).
Iba-1+ cell body area in OPL and IPL. Soma size in OHT eyes was significantly greater at 1 d than at 5, 8 or 15 d (all p < 0.01; Fig. 2f,g). Soma size in the OPL decreased marginally from 1 to 3 d, after which it decreased significantly, such that it was significantly smaller at 8 and 15 d than at 3 d (both p < 0.01; Fig. 2f). Similar results were observed in the IPL, where size was significantly smaller at 8 d than at 3 d (p < 0.05; Fig. 2g). In contrast, soma size in contralateral eyes was greatest at 3 d, after which it decreased such that it was significantly smaller at 8 and 15 d than at 3 d (both p < 0.05; Fig. 2f,g).
Arbor area of the Iba-1+ cells in the OPL and IPL. In OHT eyes, Iba-1+ arbor area in the OPL decreased marginally from 1 d to 3 d, after which it increased, such that the area was significantly greater at 8 d than at 3 d (p < 0.01; Fig. 3a) and greater at 8 d than at 5 d (p < 0.05; Fig. 3a). Arbor area in the IPL decreased significantly from 1 d to 5 d, such that the area at 3 and 5 d was significantly smaller than at 1 d (both p < 0.01; Fig. 3b). Area increased from 5 d, such that area at 8 and 15 d was significantly greater than at 3 or 5 d (all p < 0.01; Fig. 3b).
In contralateral eyes, arbor area in the OPL and IPL was smaller at 3 d than at 1 d, although this decrease was significant only in the IPL (p < 0.05; Fig. 3b). From 3 d onward, arbor area in OPL and IPL did not change significantly (Fig. 3a,b).
Iba-1+ vertical processes between OPL and OS. In OHT eyes, the number of vertical processes increased progressively from 1 to 15 d: the numbers at 5 and 15 d were greater than at 1 d (both p < 0.01; Fig. 3c), the numbers at 5 and 15 d were greater than at 3 d (p < 0.05 and p < 0.01, respectively; Fig. 3c), and the number at 15 d was greater than at 8 d (p < 0.05; Fig. 3c). In contrast, the number of vertical processes in contralateral eyes did not change significantly between 1 and 15 d (Fig. 3c).
Changes in the Iba-1+ cell in the nerve fiber layer-ganglion cell layer (NFL-GCL). The changes suffered in Iba-1+ cells of this layer after laser-induced OHT made it difficult to quantify their number. We quantified the area of the retina occupied by Iba-1+ cells (abbreviated Iba1-RA) as in previous studies 9,32,33 and we determined Iba-1+ soma area.
Comparison of OHT and contralateral eyes with naïve eyes. OHT and contralateral eyes showed several significant differences from naïve eyes. Iba1-RA was significantly greater in OHT eyes than in naïve eyes at all time points (p < 0.01; Supplementary  www.nature.com/scientificreports www.nature.com/scientificreports/ In OHT eyes, no differences were found in Iba1-RA among the superior, inferior, nasal and temporal retinal zones, either when the analysis was performed over all retinal areas or over the areas nearest to the optic disc.
Soma size was significantly greater in OHT eyes than in naïve eyes at all time points (all p < 0.01; Supplementary
Contralateral eyes. In contralateral eyes, as in naïve ones (Fig. 4), Iba-1+ cells were intensely labelled with anti-P2RY12 antibody (Fig. 9), while dendritic-like cells (Fig. 9g-i) and perivascular cells did not show staining.  Other studies have relied on rat glaucoma models, in which OHT is induced differently than in our model 21,39-45 . In the healthy retina, microglial cells remain in a quiescent state characterized by a ramified morphology. In the presence of damage, such as OHT, they become activated, their processes shorten, and their soma enlarges. www.nature.com/scientificreports www.nature.com/scientificreports/ In their state of high activation, they take on amoeboid morphology and behavior, acting as macrophages to engulf debris 9,16,32,38,46,47 . The abundance of microglia increases upon activation as a result of migration and/or proliferation 9,32,36,40,[48][49][50][51][52] . All these signs of microglial activation were observed in the present study, in OHT eyes and the contralateral normotensive eyes. We are confident that microglial activation was triggered by the increase in IOP and not by an inflammatory response to laser treatment because, in our previous work with the same OHT model 48 , no morphological signs of microglial activation were detected in animals exposed to laser in the non-draining portion of the sclera, and these eyes showed similar IOP as naïve ones. www.nature.com/scientificreports www.nature.com/scientificreports/ Laser induction of OHT in the present study considerably increased IOP at 1 d later. This increase remained at 3 d and then gradually returned to the same pressure as in contralateral eyes at 8 d. They were consistent with several previous studies using the same model 53,54 .
To label microglial cells, we used anti-Iba-1 antibody, which can reveal morphological features of activated microglial cells 55 . This antibody is expressed in the cells of monocytic lineages, including microglia, dendritic cells, monocytes and macrophages. Thus, it cannot distinguish activated microglia from infiltrating macrophages. In order to determine whether Iba-1+ cells in our study were mainly microglial cells, we stained tissue with antibody against P2RY12, which is expressed by microglia but not by monocytes, infiltrating macrophages or dendritic cells 27,56 .
Resting microglia express abundant P2RY12, which is up-regulated after injury by ATP released from damaged cells. The P2RY12 triggers the extension of microglial process towards lesions. A few hours later, P2RY12 expression is down-regulated 22 . In fact, down-regulation of P2RY12 mRNA and protein levels is one of the most sensitive markers for the transition from resting to activated microglia 26 . In our study, in naïve retinas, all Iba-1+ cells expressed P2RY12, with the exception of perivascular and dendritic-like cells. At 24 h after OHT induction, most Iba-1+ cells in the OS, plexiform layers and NFL-GCL showed abundant P2RY12 expression. This expression was down-regulated at 3 and 5 d, then it rose weakly at 8 d, reaching values similar to those of naïve eyes at 15 d. Consistent with previous studies, our results suggest that the P2RY12 receptor is down-regulated at 3 and 5 d, when the inflammation in our model was strongest. www.nature.com/scientificreports www.nature.com/scientificreports/ As expected, neither perivascular nor dendritic Iba-1+ cells showed P2RY12 inmunostaining in our study. Similarly, the sparse rounded cells observed at 1 and 15 d in the NFL-GCL and OS did not show P2RY12 expression, so they may be macrophages or monocytes infiltrating the bloodstream. P2RY12 is selectively expressed in resident microglia, and its expression correlates strongly with a ramified microglial morphology 29 . Most Iba-1+ cells with a ramified and macrophagic appearance were P2RY12+. The few ramified and macrophagic P2RY12cells were located primarily in the IPL and NFL-GCL and may derive from bloodstream-infiltrating macrophages.
When we analyzed the morphological signs of activation, we found that in the OS, the number of Iba-1+ cells was much higher in OHT and contralateral eyes than in naïve eyes at all time points after 1 d. In OHT eyes, the number increased progressively from 1 until 5 d, when it peaked, after which the number decreased. The increase in cell number in the OS may be related to the increase in IOP. Microglial activation promotes a pro-inflammatory environment that can affect retinal pigment epithelium morphology and function 57,58 . Activated Iba-1+ cells in the OS produce pro-inflammatory factors and chemokines that can trigger rupture of the external blood-retina barrier, inducing blood-cell recruitment 59 . In a study using the same OHT model as the present study, cells possibly coming from the bloodstream in the OS were observed at 15 d 32 .
In the plexiform layers, somata of Iba-1+ cells were significantly larger and arbor area significantly smaller in contralateral and OHT eyes than in naïve eyes at all time points after OHT induction. These effects were more pronounced in OHT than in contralateral eyes. In OHT, somata of Iba-1+ cells began to enlarge and processes to shorten beyond 1 d; retraction was maximum at 3 and 5 d, and somata were still large, although smaller than at 1 d. The number of Iba-1+ cells peaked at these times. From these times through the end of the 15 d, soma area decreased slightly, arbor area increased and the number of Iba-1+ cells fell, though it remained higher than in naïve eyes. Therefore, microglial activation in the plexiform layers of hypertensive eyes was greater at 3 and 5 d after OHT induction, when IOP was highest (1, 3 and 5 d).
OHT-induced microglial activation may be driven in part by the expression of growth factors, some of which are secreted by activated macroglia 52 . In a previous study, using the same experimental model as the present work, we observed activation of astrocytes and Müller glia 9 , which may have contributed to the increase in number of Iba-1+ cells in plexiform layers in OHT eyes in the current work. Previous studies using rodent models of laser-induced OHT have shown functional alterations in full-field electroretinogram recordings [60][61][62] . Indeed, the amplitude of the positive scotopic threshold response associated with retinal ganglion cell activity was significantly lower in OHT eyes than in naïve eyes throughout the 56-d measurement period. This difference was associated with a significant reduction in the a-and b-waves of the electroretinogram 60,61 . Moreover, substantial morphological alterations of the outer and inner retinal layers have been observed in the laser-induced OHT model, when rodents were analyzed for longer periods 61,62 . Thus, it is possible that neuronal damage in these layers may help explain the microglial activation observed in our study in the plexiform layers and OS.
In our study, soma area and Iba1-RA were higher in OHT and contralateral eyes than in naïve eyes in the NFL-GCL at all time points analyzed. These effects were more pronounced in OHT than contralateral eyes. In OHT eyes, Iba1-RA increased progressively from 1 d until its peak at 5 d, after which it decreased progressively until 15 d. This increase in Iba1-RA in the NFL-GCL may be explained by the increase in soma area of Iba-1+ cells. It may also be explained by entry of other cell types from the bloodstream, as a result of rupture of the internal blood-retina barrier due to the increased IOP 31,32,41 . A third possible explanation is migration of microglia from the IPL 31 .
Increased IOP triggers retinal ganglion cell degeneration and death 53,60 , which may have contributed to the microglial activation in the NFL-GCL in our study. Damaged neurons release nucleotides that can up-regulate purine receptors in microglia, triggering phagocytosis and migration [63][64][65] . In addition, deficiencies in the signaling between microglia and retinal ganglion cells activate microglia 66 . Dying neurons release HMGB1, which stimulates the NF-κβ pathway, and NADPH oxidase in microglial cells, thereby inducing the production of multiple inflammatory and neurotoxic factors 67 . A study using the same experimental model as the present work showed a lack of retrograde axonal transport in approximately 75% of the original retinal ganglion cells at 8 d after laser-induced OHT, being focal and diffuse within the retina 53 . In order to ascertain whether the number of Iba-1+ cells depends on sector-specific damage of the retinal ganglion cells, we compared the numbers of Iba-1+ cells by retinal zone (superior, inferior, nasal and temporal) in OHT eyes, over all retinal areas and over areas nearest to the optic disc. No significant sector-specific differences were observed in either case. The generalized increase in the number of Iba-1+ cells across the various retinal zones may be related to the diffuse loss of ganglion cells observed at 8 days after laser induction in this experimental model 60 .
In previous work, we demonstrated microglial activation in the contralateral eye at 1 d 31 and 15 d 32 after laser induction of OHT. Here we extend those findings to show that the activation observed at 1 d persisted throughout the full 15 d and that it was independent of IOP elevation, since the pressure was similar in contralateral and naïve eyes. Several findings point to microglial activation in contralateral eyes: (i) more abundant Iba-1+ cells in contralateral than naïve eyes in the OS from 3 d onwards; (ii) larger soma size in contralateral eyes than naïve eyes from 1 d onwards, and smaller arbor area from 3 d in the plexiform layers; (iii) greater Iba1-RA and soma area in contralateral than naïve eyes in the NFL-GCL from 1 d onwards; and (iv) more abundant microglial vertical processes between OPL and OS in contralateral than naïve eyes from 1 d onwards.
Contralateral and naïve eyes showed a similar expression of P2RY12 in all retinal layers and at all time points. These eyes did not show the P2RY12 down-regulation at 3 and 5 d observed in OHT eyes.
Microglial activation in contralateral eyes was less intense than in OHT eyes, and it remained stable over time. In OHT eyes at 1 d, enlargement of somata reached a peak, and processes began to retract. In contralateral eyes, microglia activation was maximal at 3 d when measured in terms of soma area, arbor area, and cell number in OS; or at 5 d when measured in terms of Iba1-RA. These results indicate that microglia in contralateral eyes were activated slightly later than those in OHT eyes. The processes underlying microglial activation in contralateral eyes remain unclear. Such activation does not appear to require neuronal death 9,32,48,60,68 , and it may involve an (2020) 10:4890 | https://doi.org/10.1038/s41598-020-61848-9 www.nature.com/scientificreports www.nature.com/scientificreports/ immune response triggered by impairment of the blood-retina barrier in OHT eyes 31,68 . This may help explain the observed slight delay in microglial activation in contralateral eyes.
In the present studies we have not examined the correlation of microglia activation with RGC loss. However, in a series of previous studies characterizing the model, using the same OHT-induction model and albino Swiss mice it was documented that RGC loss occurred only in the ocular hypertensive eye and not in the contralateral eye 54,60 . In the ocular hypertensive eyes, by 8 days (n = 12) after OHT-induction the population of RGCs retrogradely labelled from their main target in the brain, the superior colliculi, with of OHSt (a small molecule with similar properties to fluorogold), was reduced to approximately 25% of their fellow contralateral population; a loss that did not progressed with further survival intervals of 17 (n = 13), 35 (n = 21), or 63 (n = 13) days) 60 . When the expression of Brn3a was used as a marker to identify surviving RGCs, irrespective or their retrograde axonal transport capabilities, it was shown that by 8 days there was a loss of approximately 50% of the RGC population. Moreover, the study of the spatial distribution of retrogradely labelled RGCs with isodensity maps documented that this loss of RGCs appeared focal in wedge-shaped sectors with their vertex oriented towards the optic nerve head and their base toward retinal periphery as well as diffuse affecting the whole retina 54,60 . Thus, the main loss of RGCs occurs within the first week after OHT-induction and this is the period of time in which iba1+ cells show a peak of activation at 3 and 5d and then activation drops at 8d.
In conclusion, we report the first comparative quantification of diverse signs of microglial activation at different time points after unilateral laser-induced OHT. We examined this activation in OHT and contralateral normotensive eyes. These two eye populations were also compared with naïve eyes to allow rigorous assessment of microglial activation. We observed that microglial cells in OHT and contralateral eyes showed signs of activation at all times after laser induction, and that the number of microglia increased from 3 d onwards. In contralateral eyes microglial activation occurred slightly later, was less intense and more constant than in OHT eyes. Contralateral activation may be trigged by immune signals derived from OHT eyes. Microglial activation was greater in OHT and contralateral eyes at 3 and 5 d, coinciding with high pressure and P2RY12 down-regulation in OHT eyes. These time points of greater inflammation in this glaucoma model may be useful for identifying factors implicated in OHT-induced inflammatory processes.

Material and Methods
Animals and anesthetics. In this study adult albino mice (of Swiss strain) of 12-16 weeks of age and 40-45 g of weight were used. The animals, supplied from the breeding colony of the University of Murcia, were maintained in cages with controlled temperature and light (12-h light/dark cycle and 9-24 lux). All mice had ad libitum access to water and food. Induction of ocular hypertension (OHT) and measurement of intraocular pressure (IOP) were done under general anesthesia, using an intraperitoneal (ip) injection of a mixture of xylazine (10 mg/ kg; Rompún ® , Bayer, Barcelona, Spain) and ketamine (75 mg/kg; Ketolar ® , Parke-Davies, Barcelona, Spain). At the time of recovery after anesthesia, a tobramycin ointment (Tobrex ® ; Alcon, Barcelona, Spain) was applied to the cornea to prevent infection and drying. At all times, attempts were made to reduce pain and discomfort in the animals after surgery. Finally, the animals were sacrificed with an ip overdose of pentobarbital (Dolethal Vetoquinol ® , Veterinary Specialties, Alcobendas, Madrid, Spain). All experiments were made in accordance with Spanish law and the Guidelines for Humane Endpoints for Animals Used in Biomedical Research. In addition, the study was accepted by the Ethics Committee for Animal Research of Murcia University (Murcia, Spain) and the Animal Health Service of the Murcia Regional Ministry of Agriculture and Water (approval ID number: A1310110807). Animal procedures were made using the institutional guidelines, European Union regulations for the use of animals in research, and the Association for Research in Vision and Ophthalmology (ARVO) statement for the use of animals in ophthalmic and vision research.
Experimental groups. Mice were allocated into six groups, each containing 8 animals: one group was composed of naïve age-matched control (naïve) and five experimental groups were treated with laser in the left eye and examined at 1, 3, 5, 8 or 15 d.

Laser treatment and measurement of IOP.
To induce OHT, we treated the left eyes of the mice with a single diode laser session (Viridis Ophthalmic Photocoagulator-532 nm, Quantel Medical, Clermont-Ferrand, France) using the methods described previously 53,61 . Laser beam was applied directly on episcleral and limbal veins, and 55-76 laser burns were performed. The laser parameters were: spot size, 50-100 μm; duration, 0.5 seconds; and power, 0.3 W. IOP reading was performed on the contralateral and laser treated eyes using a rebound tonometer (Tono-Lab, Tiolat, OY, Helsinki, Finland) 40,69 . The intraocular pressure (IOP) was measured in all animals. In the five laser-treated groups, IOP was measured before, and at 1, 3, 5, and 8 d after OHT-induction. In naïve eyes IOP was taken before the animal sacrifice. Six consecutive IOP measurements were taken and then an average was made. IOP was recorded at the same time (approximately 9 a.m.), and just after anesthetizing the animal, to decrease pressure variations due to the animal's circadian rhythm 70

or spontaneous increases in the IOP 49
Immunohistochemistry. The animals were fixed by transcardiac perfusion. A saline solution was introduced briefly through the aorta and then the fixation solution composed of 4% paraformaldehyde in 0.1 M phosphate buffer (PB, pH 7.4). To maintain the orientation of the eyeball, a stitch was given at the posterior pole. Additional benchmarks such as nasal caruncle and rectus muscles were also used 71 . Subsequently the eyes were post-fixed in 4% paraformaldehyde in 0.1 M phosphate buffer (PB, pH 7.4) for two hours. Then the retinas were separated and the vitreous humor was eliminated to perform retinal whole-mounts 13,72 .
Two negative controls were also performed. In the first, there was no primary antibody, and the samples were incubated with the secondary antibody and the diluents used for the antibodies. In the second, the tissue was only incubated with the diluents used in the primary and secondary antibodies. This second control allowed assessment of how much endogenous fluorescence contributed to the observed fluorescence 74 .
Retinas were studied using a fluorescence microscope (Axioplan 2 Imaging Microscope, Carl Zeiss) to which an ApoTome device (Carl Zeiss, Munich, Germany) and a high-resolution digital camera (Cool-SNAP Photometrics, Tucson, AZ, USA) were associated. The microscope has a Zeiss 10 filter set for Alexa Fluor 488 and a Zeiss 64 filter set for Alexa Fluor 594. Using the motorized xyz microscope stage, the retinal whole-mounts were studied along the xyz axes. Cellular components located in the same xz plane were considered to be located in the same focal plane. Images were taken with the apotome device, which creates optical sections of fluorescent samples-free of scattered light and increases the resolution in Z direction compared to conventional fluorescence microscopy. Z-stacks were analyzed in Axiovision Quantification was carried in all naïve control eyes (n = 6) as well as all OHT and contralateral eyes of all animals in all the laser-treated groups as we described previously 32,65,75,76 . For the retinal count of Iba-1+ cells, in the retinal whole mounts, equivalent areas were photographed in the vertical and horizontal meridians, which crossed the optic nerve, and which included the upper, lower, nasal and temporal areas. To prevent a portion of the retinal whole-mount were being duplicated or omitted, all the fields analyzed were contiguous. Each meridian was analyzed using the motorized stage of the microscope to scan its entire length along the X-Y axis, giving an approximate total of 550 fields evaluated 32,75,76 . The photographs were made at 20x magnification, providing an area of 0.1502 mm 2 per field. Besides, to quantify Iba-1+ cells in the OPL, IPL, and NFL-GCL, the entire preparation was studied along the Z axis in depth every 2 μm.
The method used for the Iba-1+ cell quantification depended on the retinal layer. In the plexiform layers, microglial cells were distributed throughout the retina forming a regular mosaic-shaped plexus, and therefore, these cells could be individually identified and counted automatically 75,77 . However, in the NFL-GCL the microglial cells cannot be so easily identified, especially in the HTO eyes, so in this layer the Iba1-RA was used for quantification 75,77 . In OS, Iba-1+ cells easily individualized, but did not form a regular plexus, which hindered their automatic counting, therefore, in this layer a manual counting was used.
Iba-1+ cell number in the OS. In OS, the Iba-1 + cell count was performed using the "Interactive" manual counting tool in AxioVision 4.8.2 software (Carl Zeiss) that is associated with the Apotome device and the fluorescence microscope.
Iba-1+ cell number in the OPL, IPL and NFL-GCL. In these retinal layers Iba-1+ cells were counted using an algorithm developed in MATLAB by our group 75-77 that allows quantifying the number of Iba-1 + cells in the OPL and IPL and the Iba-1RA in the GCL-NFL. To quantify the number of Iba-1 + cells, we proceeded according to our protocol 75,77 . In summary, a Z-stack projection was performed on the selected images. The image was normalized to the pixel that had a higher value in the image, so that the image values were in a range between 0-1. The threshold was then imposed so that all values <0.2 were set to 0 and the remaining values were maintained. The remaining image was divided into segments and the center of mass of each segment was determined to identify the presence or absence of cells. To avoid counting twice the same cell that is between two adjacent segments, the minimum distance that separates the cells was specified. All points that are at a distance less than this minimum distance were considered to belong to the same cell, and therefore were counted only once.
For the quantification of Iba1-RA, we use the threshold tool in MATLAB in the selected images 75,77 . The thresholds determine the objects of interest by grayscale values and allow them to be differentiated from other areas of the image. After that, we quantify the Iba1-RA using the "count NFL" algorithm.
Arbor area of Iba-1+ cells in the OPL and IPL. We studied four equivalent areas in each plexiform layer along the horizontal and vertical meridians that cross the optic nerve. To have a representation of the all retinal tissue, we select different retinal areas at various distances from the optic disc, in the different quadrants of the (2020) 10:4890 | https://doi.org/10.1038/s41598-020-61848-9 www.nature.com/scientificreports www.nature.com/scientificreports/ retina as follow: the retinal area nearby to the optic disc in the superior retina; the area at two levels of eccentricity in the inferior retina; the area at three levels of eccentricity in the nasal retina and the area at four levels of eccentricity in the temporal retina. In photographs taken at 20×, the arbor area of the Iba-1 + cells in the plexiform layers was quantified 32 . For it, a polygon was drawn that connected the most distal points of the Iba-1 + cellular processes using the" interactive Measurement" tool of the AxioVision software (Carl Zeiss).

Number of vertical processes of Iba-1+ cells connecting the OPL and OS.
To make this quantification, we took 20x photographs in the retinal plane between the OPL and the OS. The vertical processes of the Iba-1+ cells that connected both retinal layers were observed as points, which were quantified using the manual counting tool of the AxioVision software.
Cell body area of Iba-1+ cells in the OPL, IPL and NFL-GCL. The cell body area was measured in the OPL, IPL and NFL-GCL using the same areas as those used for the arbor area. In photographs taken at 20×, the contours of the cell bodies were traced using the "interactive measurement" tool of the AxioVision software.
Mean P2RY12 intensity expression. In the retinal whole mounts labelled with P2RY12 antibody, equivalent areas were photographed at 20x in the vertical and horizontal meridians, including the upper, lower, nasal and temporal areas in different retinal layers (OPL, IPL and NFL-GCL). The photographs were taken with equal values of time exposure and intensity of excitation. To quantify P2RY12 expression intensity, we used an algorithm developed in MATLAB (© MathWorks, Inc) and AxioVision 4.8.2 software (Carl Zeiss) that is associated with the Apotome device and the fluorescence microscope. The algorithm allows us to identify different intensity levels of P2RY12 expression in the OPL, IPL and NFL-GCL in the different groups of study. Image Average Intensity was calculated by adding the pixel values of the green channel (P2RY12 immunostaining) of each image and dividing this result by the total number of pixels of the image. The percentages we show in this paper are the result of normalizing the Image Average Intensity values by 255 that is the maximum value each pixel can represent.
Statistical analysis. The SPSS 22 program (IBM, Chicago, IL, USA) was used to perform the statistics. Data were provided as mean ± SD. In the following parameters: (i) IOP; (ii) Iba-1+ cell number (OS, OPL and IPL); (iii) Iba1-RA (NFL-GCL); (iv) arbor area of Iba-1+ cells (OPL and IPL); (v) number of Iba-1+ vertical processes; (vi) cell body area of Iba-1+ cells (OPL, IPL and NFL-GCL), and vii) mean P2RY12 intensity expression, differences between naïve, OHT and contralateral eyes were analyzed using the Wilcoxon W test (for paired data) and Mann Whitney U test (for unpaired data). At the different time points, differences between contralateral and OHT eyes were evaluated using the ANOVA test with Bonferroni correction.
In the following parameters of OHT eyes: (i) number of Iba-1+ cells (OS, OPL and IPL) in all retinal zones; (ii) number of Iba-1+ cells (OS, OPL and IPL), taking into account only the zones nearest to the optic disc; (iii) Iba1-RA (NFL-GCL) in all retinal zones; and (iv) Iba1-RA (NFL-GCL), taking into account only the zones closest to the optic disc, the differences between several areas of the retina (superior, inferior, nasal and temporal) were analyzed using the ANOVA test with Bonferroni correction.
We consider that the differences were significant when p < 0.05.